dermatological conditions in intensive care: a secondary

Available online http://ccforum.com/content/12/S1/S1

Open AccessVol 12 Suppl 1Research articleDermatological conditions in intensive care: a secondary analysis of the Intensive Care National Audit & Research Centre (ICNARC) Case Mix Programme DatabaseSusannah MC George1, David A Harrison2, Catherine A Welch2, Kathleen M Nolan3 and Peter S Friedmann4

1Clinical Teaching Fellow in Specialties (Dermatology), St Helier Hospital, Wrythe Lane, Carshalton, Surrey SM5 1AA, UK2Intensive Care National Audit & Research Centre, Tavistock House, Tavistock Square, London WC1H 9HR, UK3Critical Care Directorate, Southampton General Hospital, Tremona Road, Southampton SO16 6YD, UK4University of Southampton, Dermatopharmacology Unit, Southampton General Hospital, Tremona Road, Southampton SO16 6YD, UK

Corresponding author: Susannah MC George, [email protected]

Received: 14 Jun 2007 Revisions requested: 14 Jun 2007 Revisions received: 27 Sep 2007 Accepted: 18 Jan 2008 Published: 18 Jan 2008

Critical Care 2008, 12:S1 (doi:10.1186/cc6141)This article is online at: http://ccforum.com/content/12/S1/S1© 2008 George et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Introduction Dermatology is usually thought of as an outpatientspecialty with low mortality, however some skin conditionsrequire intensive care. These conditions are relatively rare andhence are best studied using clinical databases or diseaseregistries. We interrogated a large, high-quality clinical databasefrom a national audit of adult intensive care units (ICUs), with theaim of identifying and characterising patients withdermatological conditions requiring admission to ICU.

Methods Data were extracted for 476,224 admissions to 178ICUs in England, Wales and Northern Ireland participating in theCase Mix Programme over the time period December 1995 toSeptember 2006. We identified admissions with dermatologicalconditions from the primary and secondary reasons foradmission to ICU.

Results A total of 2,245 dermatological admissions wereidentified. Conditions included infectious conditions (e.g.

cutaneous cellulitis, necrotising fasciitis), dermatologicalmalignancies, and acute skin failure (e.g. toxic epidermalnecrolysis, Stevens–Johnson syndrome and autoimmuneblistering diseases). These represent 0.47% of all ICUadmissions, or approximately 2.1 dermatological admissions perICU per year. Overall mortality was 28.1% in the ICU and 40.0%in hospital. Length of stay in intensive care was longest for thosewith acute skin failure (median 4.7 days for ICU survivors and5.1 days for ICU non-survivors).

Conclusion We have identified patients who not only requireintensive care, but also dermatological care. Such patients havehigh mortality rates and long ICU stays within the spectrum ofthe UK ICU population, similar to other acute medicalconditions. This highlights the importance of skin failure as adistinct entity comparable to other organ system failures.

IntroductionDermatology is usually thought of as an outpatient specialtywith low mortality. However, some skin conditions can be verysevere and the condition itself, or associated complications ofthe condition or treatment, may cause the patient to requireintensive care. There is little literature dealing specifically withintensive care dermatology [1,2]. Green et al. described theskin problems that may arise in a critically ill patient [1]. Dunnillet al. reported a case series of 27 dermatology patients admit-ted to a 30-bedded intensive care department at St Thomas'sHospital over a 14-month period [2].

Conditions that may require intensive care managementinclude toxic epidermal necrolysis (TEN), Stevens–Johnsonsyndrome (SJS), necrotising fasciitis and exfoliative dermatitis.Patients with large areas of skin involvement have problemswith fluid and electrolyte balance and with temperature controlas a result of large daily percutaneous fluid losses [2]. In criti-cally ill patients, hypoalbuminaemia is typically caused by thesystemic inflammatory response syndrome and redistribution.However, in patients with extensive skin loss, hypoalbuminae-mia also occurs as a result of cutaneous losses, hypercata-bolic state and decreased synthesis [2]. Patients with

of 10(page number not for citation purposes)

http://ccforum.com/content/12/S1/S1

http://creativecommons.org/licenses/by/2.0

http://www.biomedcentral.com/info/about/charter/

Critical Care Vol 12 Suppl 1 George et al.

impaired barrier function are also vulnerable to infection. In1991, Irvine defined the concept of skin failure and proposedthat it was a real entity comparable to any other major organsystem dysfunction [3]. However, the concept of skin failure isnot well known [4], and skin conditions that may require inten-sive care are relatively rare. For example, the incidence of TENis reported to be between 0.4 and 1.3 cases per million per-son years and that of SJS between 1–7 cases per million per-son years (which equate to approximately 50 and 240 casesper year, respectively, in the UK) [5,6]. The incidence of necr-otising fasciitis is estimated at 500 new cases per year in theUK [7]. This makes it difficult for individual clinicians to gainexperience of managing such cases. Conditions with such alow incidence are best studied using large high quality clinicaldatabases arising from multicentre audits or disease registries[8-10]. The collective experience of managing these rarecases can then be collated.

Our aim was to describe the frequency, physiological parame-ters, length of stay, and mortality of patients with dermatologi-cal conditions admitted to UK intensive care units (ICUs)participating in a national audit of intensive care over a 10-yearperiod.

Materials and methodsThe Case Mix Programme DatabaseThe Case Mix Programme (CMP) is the national, comparativeaudit of adult, general critical care units (ICUs and combinedintensive care and high dependency units) in England, Walesand Northern Ireland. Data are extracted by trained data col-lectors and undergo extensive local and central validation priorto inclusion in the CMP Database [11]. Validated data for alladmissions to units participating in the CMP between Decem-ber 1995 and September 2006 were extracted from the CMPDatabase.

The CMP has received approval from the Patient InformationAdvisory Group (PIAG), under Section 60 of the Health andSocial Care Act 2001, for the collection and use of patientidentifiable information without individual patient consent(approval no. PIAG 2-10(f)/2005).

Selection of casesIn the CMP, reasons for admission to the ICU are recordedusing a hierarchical coding method, the Intensive CareNational Audit & Research Centre (ICNARC) coding method[12]. Cases were selected if the primary or secondary reasonfor their admission to ICU was recorded as any of the followingconditions, coded within the dermatological system of theICNARC coding method:

• necrotising fasciitis;

• cutaneous cellulitis;

• orbital cellulitis;

• cutaneous melanoma;

• basal cell carcinoma;

• toxic epidermal necrolysis;

• Stevens–Johnson syndrome;

• erythema multiforme;

• exfoliative dermatitis;

• pemphigus vulgaris;

• psoriasis and pustular psoriasis;

• scleroderma.

If both the primary and secondary reasons for admission weredermatological, then the patient was categorised according tothe primary reason for admission. In addition, patients wereidentified if their primary or secondary reason for admissionwas partially coded within the dermatological system of thehierarchical coding method. Partial codes are used when aspecific condition cannot be identified within the codingmethod. Partially-coded reasons were reclassified, where pos-sible, using the process tier of the coding method (e.g. infec-tion or malignancy) and the free text field.

AnalysesAn analysis plan was agreed a priori. For each of the individualconditions listed above, we summarised the number of admis-sions and the numbers of these dying in the admitting ICU,transferred to another ICU, and dying before ultimate dis-charge from an acute hospital.

The following three groups of conditions (as classified in Table1) were analysed separately and in more detail:

1. Infective conditions;

2. Dermatological malignancies;

3. Conditions causing acute skin failure.

For these three subgroups, the following measures of casemix, activity and outcome were summarised.

Case mixThe case mix of admissions was described by the age, sex,surgical status, severe conditions in the past medical history,the Acute Physiology And Chronic Health Evaluation(APACHE) II score and predicted mortality [13], and the ICN-



Table 1

Dermatological conditions in the Case Mix Programme Database

Condition ICU admissions, n (%*) ICU deaths, n (%) ICU transfers, n (%) Ultimate hospital deaths, n (%)

Infective conditions:

Necrotising fasciitis 1,133 (0.24) 336 (29.7) 56 (4.9) 438 (41.6)

Cutaneous cellulitis 658 (0.14) 193 (29.3) 14 (2.1) 274 (42.6)

Orbital cellulitis 48 (0.01) 8 (16.7) 1 (2.1) 12 (26.7)

Wound infection† 28 (0.01) 2 (7.1) 0 (0.0) 10 (35.7)

Infected ulcer† 27 (0.01) 10 (37.0) 0 (0.0) 17 (65.4)

Abscess† 23 (< 0.01) 2 (8.7) 1 (4.3) 4 (18.2)

Gangrene† 6 (< 0.01) 1 (16.7) 0 (0.0) 1 (16.7)

Infected eczema† 3 (< 0.01) 1 (33.3) 0 (0.0) 1 (33.3)

Dermatological malignancies:

Cutaneous melanoma 80 (0.02) 6 (7.5) 0 (0.0) 13 (16.9)

Basal cell carcinoma 96 (0.02) 6 (6.3) 0 (0.0) 11 (12.0)

Squamous cell carcinoma† 15 (< 0.01) 0 (0.0) 0 (0.0) 0 (0.0)

Acute skin failure:

Toxic epidermal necrolysis 86 (0.02) 32 (37.2) 5 (5.8) 43 (50.6)

Stevens–Johnson syndrome 46 (0.01) 14 (30.4) 5 (10.9) 19 (46.3)

Erythema multiforme 19 (< 0.01) 3 (15.8) 2 (10.5) 6 (31.6)

Psoriasis and pustular psoriasis 19 (< 0.01) 8 (42.1) 0 (0.0) 8 (42.1)

Exfoliative dermatitis 16 (< 0.01) 7 (43.8) 0 (0.0) 7 (43.8)

Pemphigus vulgaris 9 (< 0.01) 4 (44.4) 0 (0.0) 4 (44.4)

Cutaneous T cell lymphoma† 2 (< 0.01) 1 (50.0) 0 (0.0) 2 (100.0)

Staphylococcal scalded skin syndrome† 1 (< 0.01) 0 (0.0) 0 (0.0) 0 (0.0)

Epidermolysis bullosa† 1 (< 0.01) 1 (100.0) 0 (0.0) 1 (100.0)

Other:

Scleroderma 36 (0.01) 15 (41.7) 1 (2.8) 19 (54.3)

Rash due to systemic infection† 18 (< 0.01) 3 (16.7) 1 (5.6) 5 (38.5)

Pressure sores† 10 (< 0.01) 2 (20.0) 0 (0.0) 3 (33.3)

Allergic reaction† 7 (< 0.01) 0 (0.0) 0 (0.0) 0 (0.0)

Allergy testing† 5 (< 0.01) 0 (0.0) 0 (0.0) 0 (0.0)

Complication following skin surgery for non malignant condition†

4 (< 0.01) 1 (25.0) 0 (0.0) 1 (25.0)

Dermatomyositis† 2 (< 0.01) 1 (50.0) 0 (0.0) 2 (100.0)

Pyoderma gangrenosum† 2 (< 0.01) 2 (100.0) 0 (0.0) 2 (100.0)

Weber-Christian disease† 1 (< 0.01) 0 (0.0) 0 (0.0) 1 (100.0)

Behçet's syndrome† 1 (< 0.01) 0 (0.0) 0 (0.0) 0 (0.0)

Wegener's granulomatosis† 1 (< 0.01) 1 (100.0) 0 (0.0) 1 (100.0)

Hypereosinophilia† 1 (< 0.01) 1 (100.0) 0 (0.0) 1 (100.0)

Radiation necrosis† 1 (< 0.01) 0 (0.0) 0 (0.0) 0 (0.0)

Axillary hyperhidrosis (thoracic sympathectomy)†

1 (< 0.01) 0 (0.0) 0 (0.0) 0 (0.0)

Total 2,406 (0.51) 661 (27.5) 86 (3.6) 906 (39.6)

*Percentage of all admissions in the Case Mix Programme Database. †Identified from partial code and text field. ICU, intensive care unit.



ARC physiology score and predicted mortality [14]. In addi-tion, for admissions with TEN and related conditions (SJS anderythema multiforme), the disease-specific SCORTEN severityscoring system was also used. The SCORTEN comprises ascore for seven factors relating to mortality: age greater than40 years; heart rate greater than 120 min-1; cancer; involvedbody area greater than 10%; serum urea greater than 10mmol.l-1; serum bicarbonate less than 20 mmol.l-1; and serumglucose greater than 14 mmol.l-1 [15]. The APACHE II, ICN-ARC and SCORTEN models were calculated using computeralgorithms based on raw data recorded from the first 24 h inICU.

The most frequently occurring non-dermatological reasons foradmission recorded in the database were also reported for allpatients with dermatological conditions.

OutcomeOutcome was described by the mortality in the original admit-ting ICU and at ultimate discharge from an acute hospital. Foradmissions with TEN and related conditions, hospital mortalitywas also reported by SCORTEN. The ability of SCORTEN todiscriminate between hospital survivors and non-survivors wasevaluated by the area under the receiver operating character-istic (ROC) curve [16] and compared with the discriminationof the APACHE II score and the ICNARC risk predictionmodel.

The ROC curve is a plot of the sensitivity against the specifi-city of the score for predicting death at each different value.The area under the curve, also called the concordance or c-index, is equivalent to the probability that a randomly selectednon-survivor will have a higher score than a randomly selectedsurvivor.

ActivityActivity was described by the number of patients transferredto another ICU, the number transferred for more specialisedcare, the length of stay in ICU by ICU discharge status (dis-charged, transferred or died), and the total length of stay in anacute hospital by ultimate hospital survival status.

ResultsOut of 476,224 admissions to 178 adult general ICUs from 1December 1995 to 30 September 2006, 2,245 were cases inwhich a dermatological condition was fully specified as the pri-mary or secondary reason for admission. A further 213 partiallycoded dermatological reasons were identified. Of these, threecould be given a full code within the coding method, 158 werecategorised based on the free text field and the remaining 52had insufficient detail recorded and were excluded from theanalyses. The resulting 2,406 dermatological admissions(Table 1) represented 0.51% of all admissions in the data-base, or approximately two dermatological admissions per unitper year.

Overall, 661 patients (27.5%) with a dermatological conditiondied in intensive care with a total of 906 patients (39.6%)dying before ultimate discharge from an acute hospital. Therewere 86 (3.6%) transfers of patients to another ICU.

The most frequent dermatological reasons for admission tointensive care were infective conditions. There were 1,133cases of necrotising fasciitis and 658 cases of cutaneous cel-lulitis, which together made up three quarters of all dermato-logical admissions to intensive care.

Excluding conditions representing fewer than 20 admissions,mortality was lowest for patients with cutaneous malignancies,with 11 (12.0%) hospital deaths amongst patients with basalcell carcinomas and highest for those with scleroderma with19 (54.3%) hospital deaths.

Patients with dermatological malignancies were older thanthose with infective conditions or acute skin failure (mean age65 years versus 58 and 52 years, respectively), and there wasa higher percentage of male patients in this group (59% ver-sus 52% and 45%) (Table 2).

Patients with infective conditions were mostly medical admis-sions (52.8%) or were admitted following emergency surgery(38.3%). By comparison, the vast majority of those with der-matological malignancies were admitted following electivesurgery (72.3%) and most of those with acute skin failure werenon-surgical admissions (85.4%) (Table 2).

Disease severity scores, APACHE II and the ICNARC physiol-ogy score, were highest in those with infective conditions andlowest in those with cutaneous malignancies.

Length of intensive care stay was longest for those with acuteskin failure (median 4.7 days for both ICU survivors and non-survivors). Length of intensive care stay and total hospital staywas shortest for those with dermatological malignancies.

The most commonly recorded non-dermatological reasons foradmission among these patients were for septic shock/septi-caemia (n = 573, 23.8%), acute renal failure (n = 168, 7.0%)and pneumonia (n = 95, 4.0%) (Table 3).

For admissions with TEN, SJS and erythema multiforme (n =145), mortality increased steeply with SCORTEN (Figure 1),rising from around 20% for scores of 0–2 to over 70% forscores of 4 or more. The area under the ROC curve forSCORTEN was 0.762 (95% confidence interval 0.685–0.838) (Figure 2). This compared with values of 0.737(0.655–0.819) for the APACHE II score (in 137 eligiblepatients), and 0.795 (0.722–0.867) for the ICNARC model,suggesting that the ICNARC model was best for discriminat-ing between survivors and non-survivors in this patient group.However, with relatively small numbers, none of the differ



Table 2

Case mix, outcome and activity for major subgroups of dermatological admissions

Infective conditions Dermatological malignancies Acute skin failure

Admissions, n (% of all admissions) 1,926 (0.40) 191 (0.04) 199 (0.04)

Age, mean (SD) 57.6 (16.5) 65.1 (15.0) 51.7 (20.8)

Male, n (%) 1,009 (52.4) 112 (58.6) 90 (45.2)

Surgical status, n (%):

Non-surgical 1,016 (52.8) 38 (19.9) 170 (85.4)

Elective 172 (8.9) 138 (72.3) 16 (8.0)

Emergency 737 (38.3) 15 (7.9) 13 (6.5)

Past medical history*, n (%):

Steroid treatment 65 (3.4) 3 (1.6) 12 (6.1)

Chemotherapy 43 (2.3) 8 (4.2) 13 (6.6)

Metastatic disease 17 (0.9) 25 (13.2) 0 (0.0)

Chronic renal replacement therapy 37 (1.9) 1 (0.5) 5 (2.5)

Severe respiratory disease 32 (1.7) 1 (0.5) 3 (1.5)

Very severe cardiovascular disease 26 (1.4) 3 (1.6) 3 (1.5)

Radiotherapy 15 (0.8) 11 (5.8) 1 (0.5)

Lymphoma 19 (1.0) 1 (0.5) 6 (3.0)

Acute myelogenous/lymphocytic leukaemia or multiply myeloma 11 (0.6) 0 (0.0) 3 (1.5)

Portal hypertension 12 (0.6) 0 (0.0) 1 (0.5)

Biopsy proven cirrhosis 12 (0.6) 0 (0.0) 0 (0.0)

Chronic myelogenous/lymphocytic leukaemia 8 (0.4) 1 (0.5) 1 (0.5)

Home ventilation 8 (0.4) 1 (0.5) 0 (0.0)

Congenital immunohumoral or cellular immune deficiency state 5 (0.3) 0 (0.0) 1 (0.5)

Hepatic encephalopathy 5 (0.3) 0 (0.0) 0 (0.0)

AIDS 2 (0.1) 0 (0.0) 0 (0.0)

APACHE II score, mean (SD) 19.1 (7.5) 14.5 (5.3) 18.1 (7.0)

ICNARC model:

Physiology score, mean (SD) 22.7 (10.4) 12.1 (6.7) 21.2 (10.6)

Predicted mortality, median (IQR) 37.6 (16.7–64.5) 10.1 (6.2–19.2) 32.7 (12.3–62.7)

ICU length of stay (days), median (IQR):

ICU survivor 4.1 (1.7–10.0) 1.0 (0.8–2.0) 4.7 (1.9–14.8)

ICU non-survivor 1.9 (0.7–5.8) 1.0 (0.3–4.0) 4.7 (2.0–10.0)



DiscussionThis report provides information about the range of dermato-logical conditions requiring intensive care in England over an11-year period (1995–2006). Of 476,224 admissions to 178ICUs in England, Wales and Northern Ireland, 2,406 dermato-logical admissions were identified. While rare, these condi-tions are important as they have a high mortality and oftenrequire long intensive care and hospital stays. Overall, 28% ofpatients died in intensive care, with a total of 40% dying inhospital. Patients with acute skin failure had the highest mor-tality with 35% dying in intensive care and a total of 47% dyingin hospital. By comparison, of the overall general adult inten-

sive care population, 20% die in the ICU and 31% die in hos-pital [11]. While the population of patients admitted tointensive care is very heterogeneous, comparisons with theaverage for all intensive care admissions allow us to place thedermatological admissions within the spectrum of all ICUadmissions. The hospital mortality of patients admitted to ICUwith acute skin failure is comparable to that of other acutemedical conditions, such as pancreatitis (42%) [17] and pneu-monia (49%) [18].

As well as having higher mortality, dermatology patients withinfective conditions and acute skin failure also have longer

Transfer to another ICU 4.0 (0.9–9.6) - 3.5 (0.7–11.6)

Total acute hospital length of stay (days), median (IQR):

Hospital survivor 39 (22–66) 17 (8–28) 34 (16–75)

Hospital non-survivor 9 (3–29) 6 (3–22) 15 (7–31)

ICU bed days (% of total bed days) 13,808 (0.60) 391 (0.02) 1,923 (0.08)

Transfer to another ICU, n (%) 72 (3.7) 0 (0.0) 12 (6.0)

ICU mortality, n (%) 553 (28.7) 12 (6.3) 70 (35.2)

Ultimate hospital mortality, n (%) 757 (41.5) 24 (13.0) 90 (46.6)

Values are n (%), mean (SD) or median (IQR), as indicated. *Percentage of admissions with evidence to assess past medical history. APACHE, Acute Physiology And Chronic Health Evaluation; ICNARC, Intensive Care National Audit & Research Centre; ICU, intensive care unit; IQR, interquartile range; SD, standard deviation.

Table 2 (Continued)

Case mix, outcome and activity for major subgroups of dermatological admissions

Table 3

Top 10 most common non-dermatological reasons for admission for patients with both dermatological and non-dermatological reasons recorded

Reason Associated non-dermatological reason for admission n (%)

1 Septic shock/septicaemia 573 (23.8)

2 Acute renal failure 168 (7.0)

3 Pneumonia 95 (4.0)

4 Chronic obstructive pulmonary disease (COPD) 36 (1.5)

5 Non-cardiogenic pulmonary oedema 31 (1.3)

6 Diabetes mellitus 27 (1.1)

7 Supra-ventricular tachycardia, atrial fibrillation or flutter 26 (1.1)

8(=) Hypovolaemic shock 25 (1.1)

8(=) Morbid obesity 25 (1.1)

10 Chronic renal failure 22 (0.9)



intensive care and hospital stays than average for the generaladult intensive care population. The median length of stay inintensive care for all adult patients is 1.7 days for survivors and2.0 days for non-survivors, whereas for dermatology patients,survivors with infective conditions and acute skin failure hadmedian ICU stays of between 4 and 5 days. In the presentstudy, ICU deaths in the acute skin failure group occurredparticularly late, with a median ICU stay of 5 days for non-sur-vivors. The median total hospital stay for patients admitted toICUs is 16 days for survivors and 9 days for non-survivors [11].Our results on hospital length of stay are consistent with twoprevious retrospective cohort studies of patients with necrotis-ing fasciitis, which reported mean hospital stays of around 31days for survivors and 12 days for non-survivors [19,20]. Oneof these studies also reported the overall mean ICU stay as 21days [20].

ences in ROC curves were statistically significant (SCORTEN versus APACHE II, Chi-squared = 0.14, p = 0.73; SCORTEN versus ICN-ARC model, Chi-squared = 1.08, p = 0.30; ICNARC model versus APACHE II, Chi-squared = 1.52, p = 0.22).Compared with the average for the general adult intensivecare population, APACHE II scores were higher for dermatol-ogy patients with infective conditions and acute skin failure(mean scores 19.2 and 18.0, respectively), but lower for der-matological malignancies (mean score 14.5). These reflecttypical scores for conditions that are predominantly emer-gency and elective, respectively. They compare with meanAPACHE II scores of 16.5 for all UK intensive care patients[11] and 19.2 for a previous cohort of 166 patients with necr-otising fasciitis [19].

The most common dermatological reason for admission tointensive care (1133 cases) was necrotising fasciitis. Over a10-year period this corresponds to an overall treated inci-dence of approximately 280 patients per year in the UK. As thetotal UK incidence is estimated at 500 new cases per year [7],

a significant proportion of these cases must be managedeither in more specialised units, or more likely in high depend-ency units or on the general ward. Mortality in previouslyreported case series of patients with necrotising soft tissueinfections varies from 17% to 33% [19-22]. Our patients withnecrotising fasciitis had an overall mortality of 41.6%. Thishigh mortality is most likely to be due to selection of only themost severe cases for treatment in an ICU due to the limitedcritical care resources in the UK.

Our study included 145 patients who had a diagnosis of TENand related conditions (SJS and erythema multiforme). Thedata available in the CMP Database allowed us to produceROC curves for SCORTEN, the APACHE II score and theICNARC model for these cases. SCORTEN was developedusing a sample of 165 patients and validated on a further 75patients from the same dermatology ICU [15]. The area underthe ROC curve in the validation sample was 0.82 (95% confi-dence interval 0.74 to 0.90), which indicates good discrimina-tory power. Further work using a large proportion of thepatients from the original study found that SCORTEN calcu-lated on any of the first 5 days of admission produced ROCvalues of over 0.8 [23]. The area under the ROC curve forSCORTEN in our study was 0.763, which was better than theAPACHE II score (0.737) but not as good as the ICNARCmodel (0.795), although these differences were not statisti-cally significant. SCORTEN is a simple model, which is quickand easy to use at the bedside. By comparison, APACHE IIand the ICNARC model are more complex and are designedto give estimates of the risk of hospital mortality across alladmissions to intensive care. The comparable performance ofthe simple SCORTEN model to these more complex modelswas therefore impressive.

Only a small number of patients had exfoliative dermatitis,pemphigus vulgaris, psoriasis and scleroderma, however all ofthese conditions had a high mortality. Even smaller numberswere identified with dermatomyositis, cutaneous T cell lym-phoma, staphylococcal scalded skin syndrome and epidermo-lysis bullosa. The rarity of these conditions makes it difficult togain experience of managing such cases.

The admission to ICU of patients with comparatively low riskconditions, such as basal cell carcinoma, may seem surprising.However, reviewing the free text field of the database for thesecases, where used, indicated that these were likely to beadmissions following complex surgery, in particular operationsinvolving facial reconstruction. This would explain the relativelylow but not insignificant mortality of 12% among this group.

Our study shows the most common non-dermatological rea-sons for admission in patients with both dermatological andnon-dermatological reasons recorded were sepsis, acuterenal failure and pneumonia. Unsurprisingly, these representcommon and direct complications of severe skin disorders.

Figure 1

Ultimate hospital mortality by SCORTEN for admissions with toxic epi-dermal necrolysis and related conditions (n = 145)Ultimate hospital mortality by SCORTEN for admissions with toxic epi-dermal necrolysis and related conditions (n = 145).



They suggest that in some cases the skin condition alone doesnot precipitate admission to ICU, but rather the associatedcomplications. In terms of pre-existing conditions, only themost severe illnesses in a patient's past medical history areroutinely recorded in the CMP Database. A review of necrotis-ing fasciitis found that pre-existing conditions included diabe-tes mellitus (incidence 21–64%), peripheral vascular disease(15–80%), intravenous drug abuse (8–77%), obesity (18–46%), chronic alcoholism (12–31%) and malnutrition (18–40%) [24]. More recent studies have found that the most fre-quent comorbidities were diabetes (19–37%) [19-21], obes-ity (17–31%) [20,21], hypertension (35%) [20] andintravenous drug abuse (30%) [19]. The low rates of condi-tions such as diabetes and obesity in our results reflect thefact that these conditions are only recorded in the CMP Data-base if they are considered to have had a significant influenceon the decision to admit the patient to intensive care.

The dermatological admissions identified therefore representa combination of those admitted for their primary dermatolog-ical condition, such as those with acute skin failure, a direct

complication of that condition, such as septic shock or renalfailure, or as a result of their treatment, such as those admittedfor monitoring following major surgery or following an acutecomplication during minor surgery. It is in the first of thesethree groups where the involvement of dermatologists may beof particular benefit in terms of diagnostic advice and knowl-edge of therapeutic options.

The high mortality and long duration of ICU stay of patientswith skin conditions compared to the general adult ICU popu-lation highlights the importance of skin failure as a cause ofmorbidity and mortality. Although conditions causing skin fail-ure are rare, it is important to recognise that this is no less seri-ous than any other organ system failure. As the broad aims ofICU management are to maintain homeostasis, disruption ofskin function results in specific management problems. Thesenecessitate a multidisciplinary approach to management withinvolvement of both intensivists and dermatologists, andmeticulous nursing care. The recent UK government proposalto move dermatology services into the community [25] mayreduce the availability of a specialist dermatological opinion

Figure 2

Receiver operating characteristic curves for SCORTEN, the APACHE II score and the ICNARC model for admissions with toxic epidermal necrolysis and related conditions (n = 145)Receiver operating characteristic curves for SCORTEN, the APACHE II score and the ICNARC model for admissions with toxic epidermal necrolysis and related conditions (n = 145).



and potentially compromise the care of this small number ofseverely ill patients.

The size of the dataset available provides information on alarge number of patients; however it also has limitations interms of lacking detail. For example, we have no information onthe accuracy of the diagnoses, whether they were made by anintensivist or a dermatologist, and whether diagnosis was con-firmed by skin biopsy. Assigning a diagnosis to such patientsis further complicated by the various diagnostic criteria thatmay have been used, for example in cases of SJS or TEN [5].In order to calculate SCORTEN we assumed that patientscoded as TEN had involvement of greater than 10% of bodysurface area and other patients did not. Patients with between10 and 30% body surface area involvement may be consid-ered 'overlap TEN/SJS' [6] and, if these were recorded solelyas SJS within our data, then the SCORTEN for these patientswould be incorrect. The available data do not provide any infor-mation on the laboratory and physiological parameters beforeadmission to intensive care, and certain prognostic factors ofparticular relevance to these patients, for example the percent-age body surface area involvement, are not recorded in thedatabase.

The conditions included within the ICNARC coding method[12], used to code the reasons for admission to ICU, are notfully comprehensive. However, as this is a hierarchical system,a method exists for coding any condition not specified in thecoding method by completing the hierarchical coding to asmuch depth as possible and then recording the actual condi-tion in the free text field. We identified 213 partially coded der-matological admissions, of which 161 could be fully classifiedusing the information in the text field. For example, squamouscell carcinomas are not included in the coding method but canbe recorded as 'surgical/dermatological/skin/tumour or malig-nancy' with the final condition of 'squamous cell carcinoma' inthe free text field. It is, however, possible that some squamouscell carcinomas were miscoded as melanomas or basal cellcarcinomas. Two squamous cell carcinomas recorded asbasal cell carcinomas were identified and reclassifiedappropriately.

This study identifies patients with dermatological conditionsrequiring admission to intensive care. However, it provides noinformation on other patients with the same conditions, whorequire inpatient care, but are not severe enough to needintensive care or those who are not considered fit enough forintensive care. For example an elderly patient with cellulitismight die from sepsis, but not have been considered to be acandidate for intensive care. Conversely, more severe casesmay have been referred directly to burns units, about which wehave no data. A number of patients with infective conditions,TEN and related conditions, and scleroderma were transferredto other ICUs. It is likely that some of these patients were

transferred to burns units where treatment outcomes may bebetter for necrotising fasciitis and TEN [21,26].

While our study considers specifically those patients whoseprimary or secondary reason for admission to intensive carewas a dermatological condition, there is also considerablescope for patients with other conditions in intensive care torequire dermatological support [1,2]. The CMP dataset hasrecently been revised to incorporate the new Critical CareMinimum Data Set [27], which includes days of dermatologi-cal support (defined as patients with major skin rashes, exfo-liation or burns, use of multiple large trauma dressings, or useof complex dressings). Once sufficient data are available, it willbe possible to examine all admissions to intensive care thatrequire such dermatological support.

ConclusionThis study highlights the importance of dermatological condi-tions in the ICU. Skin conditions necessitating ICU admissionmay be associated with severe physiological abnormalities,and such patients have higher mortality and longer intensivecare stays than average for the UK adult ICU population. Thisstudy provides further evidence for the importance of skin fail-ure as a distinct entity, comparable to other major organ sys-tem failures, with high mortality.

List of abbreviationsAPACHE = Acute Physiology And Chronic Health Evaluation;CMP = Case Mix Programme; ICNARC = Intensive CareNational Audit & Research Centre; ICU = intensive care unit;SJS = Stevens–Johnson syndrome; TEN = toxic epidermalnecrolysis.

Competing interestsThe authors declare that they have no competing interests.

Authors' contributionsSMCG and DAH conceived the study. SMCG, DAH, KMNand PSF agreed the analysis plan. DAH and CAW performed

Key messages

• Some patients with severe skin conditions require inten-sive care.

• These conditions include toxic epidermal necrolysis, Stevens–Johnson syndrome, necrotising fasciitis, exfo-liative dermatitis, and cutaneous malignancies.

• ICU admission may be the result of the skin condition alone, e.g. for acute skin failure, or following complica-tions of the condition or its treatment, e.g. following complex surgery for a malignant skin tumour.

• Patients with acute skin failure have higher mortality and longer intensive care stays than average for the UK adult ICU population.



the analyses. SMCG and DAH drafted the manuscript. Allauthors contributed to interpretation of the results and criticalrevision of the manuscript and have seen and approved thefinal version.

AcknowledgementsThis study was supported by ICNARC. The authors wish to thank every-one in the critical care units participating in the CMP [28].

References1. Green T, Manara AR, Park GR: Dermatological conditions in the

intensive care unit. Hosp Update 1989, 15:367-376.2. Dunnill MGS, Handfield-Jones SE, Treacher D, McGibbon DH:

Dermatology in the intensive care unit. Br J Dermatol 1995,132:226-235.

3. Irvine C: 'Skin failure' – a real entity: discussion paper. J R SocMed 1991, 84:412-413.

4. Inamadar AC, Palit A: Acute skin failure: concept, causes, con-sequences and care. Indian J Dermatol Venereol Leprol 2005,71:379-385.

5. Letko E, Papaliodis DN, Papaliodis GN, Daoud YJ, Ahmed AR,Foster CS: Stevens–Johnson syndrome and toxic epidermalnecrolysis: a review of the literature. Ann Allergy AsthmaImmunol 2005, 94:419-436.

6. Roujeau JC: The spectrum of Stevens–Johnson syndrome andtoxic epidermal necrolysis: a clinical classification. J InvestDermatol 1994, 102:28S-30S.

7. Hasham S, Matteucci P, Stanley PRW, Hart NB: Necrotisingfasciitis. BMJ 2005, 330:830-833.

8. Black N: Developing high quality clinical databases. BMJ1997, 315:381-382.

9. Kaufman DW: Epidemiological approaches to the study oftoxic epidermal necrolysis. J Invest Dermatol 1994,102:31S-33S.

10. Rzany B, Mockenhaupt M, Baur S, Schröder W, Stocker U, MuellerJ, Holländer N, Bruppacher R, Schöpf E: Epidemiology of ery-thema exsudativum multiforme majus, Stevens–Johnson syn-drome, and toxic epidermal necrolysis in Germany (1990–1992): structure and results of a population-based registry. JClin Epidemiol 1996, 49:769-773.

11. Harrison DA, Brady AR, Rowan K: Case mix, outcome and lengthof stay for admissions to adult, general critical care units inEngland, Wales and Northern Ireland: the Intensive CareNational Audit & Research Centre Case Mix ProgrammeDatabase. Crit Care 2004, 8:R99-R111.

12. Young JD, Goldfrad C, Rowan K: Development and testing of ahierarchical method to code the reason for admission to inten-sive care units: the ICNARC Coding Method. Br J Anaesth2001, 87:543-548.

13. Knaus WA, Draper EA, Wagner DP, Zimmerman JE: APACHE II: aseverity of disease classification system. Crit Care Med 1985,13:818-829.

14. Harrison DA, Parry GJ, Carpenter JR, Short A, Rowan K: A newrisk prediction model for critical care: the Intensive CareNational Audit & Research Centre (ICNARC) model. Crit CareMed 35:1091-1098.

15. Bastuji-Garin S, Fouchard N, Bertocchi M, Roujeau JC, Revuz J,Wolkenstein P: SCORTEN: a severity-of-illness score for toxicepidermal necrolysis. J Invest Dermatol 2000, 115:149-153.

16. Hanley JA, McNeil BJ: The meaning and use of the area undera receiver operating characteristic (ROC) curve. Radiology1982, 143:29-36.

17. Harrison DA, D'Amico G, Singer M: Case mix, outcome, andactivity for admissions to UK critical care units with severeacute pancreatitis: a secondary analysis of the ICNARC CaseMix Programme Database. Crit Care 2007, 11(Suppl 1):S1.

18. Woodhead M, Welch CA, Harrison DA, Bellingan G, Ayres JG:Community-acquired pneumonia on the intensive care unit:secondary analysis of 17,869 cases in the ICNARC Case MixProgramme Database. Crit Care 2006, 10(Suppl 2):S1.

19. Anaya DA, McMahon K, Nathens AB, Sullivan SR, Foy H, Bulger E:Predictors of mortality and limb loss in necrotizing soft tissueinfections. Arch Surg 2005, 140:151-157.

20. Endorf FW, Supple KG, Gamelli RL: The evolving characteristicsand care of necrotizing soft-tissue infections. Burns 2005,31:269-273.

21. Faucher LD, Morris SE, Edelman LS, Saffle JR: Burn center man-agement of necrotizing soft-tissue surgical infections inunburned patients. Am J Surg 2001, 182:563-569.

22. Elliott DC, Kufera MA, Myers RAM: Necrotizing soft tissue infec-tions: risk factors for mortality and strategies for management.Ann Surg 1996, 224:672-683.

23. Guégan S, Bastuji-Garin S, Poszepczynska-Guigné E, Roujeau JC,Revuz J: Performance of the SCORTEN during the first fivedays of hospitalization to predict the prognosis of epidermalnecrolysis. J Invest Dermatol 2006, 126:272-276.

24. Cunningham JD, Silver L, Rudikoff D: Necrotizing fasciitis: a pleafor early diagnosis and treatment. Mt Sinai J Med 2001,68:253-261.

25. Department of Health: Our health, our care, our say: a new direc-tion for community services London: Her Majesty's StationaryOffice; 2006.

26. Hettiaratchy S, Moloney D, Clarke J: Patients with acute skinloss: are they best managed on a burns unit? Ann R Coll SurgEngl 2001, 83:26-29.

27. UK Department of Health, Computing for Health: Critical CareMinimum Data Set. [http://www.dh.gov.uk/assetRoot/04/12/65/14/04126514.pdf].

28. Case Mix Programme participating units [http://www.icnarc.org/audit/cmp/participating-units]


http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Abstract&list_uids=7888359




















































http://www.dh.gov.uk/assetRoot/04/12/65/14/04126514.pdf

http://www.dh.gov.uk/assetRoot/04/12/65/14/04126514.pdf

http://www.icnarc.org/audit/cmp/participating-units

http://www.icnarc.org/audit/cmp/participating-units

dermatological conditions in intensive care: a secondary

Health & Medicine